Skip to main content

Full text of "Food Habits of Nesting Success of Wood Storks in Everglades National Park in 1974"

See other formats

r ji.^'i ' f^ 

Clemson University 

lillll lllillllllAill 

3 1604 019 691 569 




? ** 



SEP 8 1978 

1. ■ 

Natural Resources Report 
Number 16 

U.S. Department of the Interior 
National Park Service 

LlU'i JUU'ILL'iiiJ>r 



John C. Ogden 

Research Department 
National Audubon Society 
Tavernier , Florida 

James A. Kushlan 

U.S. National Park Service 

Everglades National Park 

Homestead, Florida 

James T. Tilmant 

U.S. National Park Service 

Biscayne National Monument 

Homestead, Florida 

U.S. Department of the Interior, National Park Service 
Natural Resources Report Number 16 • 1978 

For sale by the Superintendent of Documents, U.S. Government Printing Office 
Washington, D.C. 20402 

Stock Number 024-005-00"r;-2 


The Wood Stork (Mysteria ameriaana) is a wading bird adapted to fluctuating water levels 
such as those found in the Everglades of Everglades National Park. The population of 
Everglades Wood Storks declined through the 1960s and they did not nest successfully from 
1967 through 1973. This paper describes a study conducted in 1974 on the factors affecting 
food supplies and successful nesting of Wood Storks in Everglades National Park. Since 
Wood Storks nested successfully in 1974, this study provided the first opportunity in 7 
years to obtain information necessary for the preservation of the species in Everglades 
National Park. As a result of this initial study, we recommend areas of future research 
required for proper management of the Everglades ecosystem. 

Wood Storks began nesting in late January 1974 and the 2,000 nesting storks successfully 
reared 1,900 young. Since 1974 was a dry year, it continued the recent but unnatural 
pattern of storks nesting successfully in dry years, which began after the institution of 
water control in 1962. An important result of this year's nesting was that the time of 
nesting correlated with the rate of drying in the Everglades, substantiating a previously 
derived relation. The 1974 results proved that sufficient food can still be produced in 
the highly altered southern Florida environment to permit successful nesting of the remnant 
population of Everglades Wood Storks, although with loss of habitat and artificial impound- 
ment of water, storks were forced to fly 130 km from the colony to feed young late in the 
nesting season. 

Preliminary information was obtained on the abundance of Wood Stork prey in the southern 
Everglades and estuarine areas of the park. The first substantial understanding of food 
habits of Everglades Wood Storks also was acquired through collection of food items from 
nestling and adult storks. The diet of storks was composed almost entirely of fish and 
only a few species made up most of the food consumed. Storks fed where food was relatively 
concentrated and selectively consumed the relatively larger fish of those available. 

The rapid population decline and changes in Wood Stork nesting success of recent years 
coincided with the institution of water control. Analyzing these changes, as well as deri- 
ving the proper management to preserve the natural Everglades ecosystem are complicated by 
diversion of surface water to the western part of the southern Everglades rather than 
through the natural drainage basin. 

Cover photo Wood Stork in Typical Feeding Posture i>y Frank Mazzotti 

As the Nation's principal conservation agency, the Department of 
the Interior has responsibility for most of our nationally owned 
public lands and natural resources . This includes fostering the 
wisest use of our land and water resources, protecting our fish 
and wildlife, preserving the environmental and cultural values 
of our national parks and historical places, and providing for 
the enjoyment of life through outdoor recreation. The Department 
assesses our energy and mineral resources and works to assure 
that their development is in the best interests of all our 
people. The Department also has a major responsibility for 
American Indian reservation communities and for people who live 
in Island Territories under U.S. administration. 



This paper is a report on a study of the 
ecology of the Wood Stork {Myateria ameri- 
oana) in Everglades National Park, conduct- 
ed from September 1973 to June 1974. The 
Wood Stork is an important link in the food 
chains of the Everglades and estuarine 
ecosystems of southern Florida. It is 
adapted to the fluctuating water conditions 
which typify southern Florida (Kahl 1964) . 
Its nonvisual feeding behavior (Kahl and 
Peacock 1963) requires high prey density 
(Kahl 1964) . The rapid decrease in popula- 
tion over the past decades suggests that it 
is the wading bird species most susceptible 
to adverse alteration of natural ecological 
conditions. Because of wading birds' eco- 
logical importance, historically large 
populations, and adaptations to fluctuating 
water marslilands , understanding the ecology 
of the Wood Stork and of other less sensi- 
tive species is necessary for the manage- 
ment and preservation of the natural Ever- 
glades ecosystem. 

The breeding status of the Wood Stork in 
Everglades National Park was especially 
critical at the start of this study. The 
last previous successful nesting season was 
1967, and during the 1960s and early 1970s 
the breeding population fell from approxi- 
mately 5,000 to 2,000 birds. Thus, the 

successful 1974 nesting season provided the 
first opportunity in 7 years to gather in- 
formation necessary for the preservation of 
this species in Everglades National Park. 
The background of the current program 
was a report by Kushlan et al . (1975) which 
attempted to discern historic relationships 
between surface water changes, fish popula- 
tions, and the nesting success of Wood 
Storks in Everglades National Park. The 
results showed that before the completion 
of the levee system north of the park in 
1962, Wood Storks nested successfully in 
those years of high water levels, high sur- 
face water discharge, and high rates of 
drying late in the nesting season. Kushlan 
et al. (1975) concluded that after 1962 the 
above pattern of water fluctuation no long- 
er permitted successful nesting in the 
present, highly altered system and recog- 
nized that future management must be de- 
vised within the constraints imposed by the 
altered system. Considering information 
after 1962, Kushlan et al. (1975) found 
that nesting failed primarily because of 
desertion, which correlated with rising 
water levels, either a small water level 
rise early in the nesting season or a large 
rise at the beginning of the rainy season. 
They also found that the latter type of 

desertion resulted from late nesting and 
that, after 1962, early nesting occurred 
with rapid drying rates early in the dry 
season. They noted that since nesting by 
Wood Storks would be initiated and sus- 
tained only through adequate food supplies, 
the key to nesting success must lie in the 
production and timely availability of fish, 
about which little was known. Kushlan et 
al. (1975) hypothesized that freshwater 
fish may move long distances as water lev- 
els fell, emigrating ahead of progressively 
drying conditions and thereby increasing 
population densities and availability to 
Wood Storks. 

In the present study, we attempted to 
further investigate the findings and test 
hypotheses that reducing surface water dis- 
charge into the southern Everglades, known 
as Shark River Slough, would increase dry- 
ing rates and bring about earlier nesting 
of Wood Storks. To accomplish this, it was 

proposed that gates in the two eastern 
spillways that release water into the park 
be closed in November and December and 
water be diverted through two western spill- 
ways. Discharge through the western spill- 
ways apparently has little effect on the 
slough since it drains mostly to the south- 
west (F. A. Nix pers. comm.). 


Everglades National Park is located on 
the subtropical southern tip of Florida 
(Fig. 1) . The central feature and primary 
watershed of the park are the freshwater 
marshes of the southern Everglades, knov\Ti 
as Shark River Slough. A dominant factor 
within the Everglades ecosystem is a sea- 
sonal rainfall pattern and the resulting 
seasonal fluctuation of water levels. 
Water levels are high in summer and fall 
and drop through the winter and spring. 



FIGURE 1. Map of southern Florida. 

National Park Service 

Conservation area 

Conservation area 

+ Gaging Stations 
V Rainfall stations 
O Permanent fish trap sites 
^ Fish samples at stork 

feeding sites 
• Feeding site where stork 

A Feeding site where both 

fish sample & regurgitation 

■ Stork rookery, 197^ 
□ Unused rookery site 

FIGURE 2. Location of fish samples and other sites in southern Florida. 

The Everglades is composed primarily of 
sawgrass marsh and marsh prairie. Water 
from the Everglades flows into the estuar- 
ies where mangrove swamps and coastal flats 
are the predominant habitats. The natural 
Everglades extended from Lake Okeechobee 
to the mangrove swamps of the southwestern 
coast. In the early 1960s the northern 

Everglades was impounded by levees to form 
three Water Conservation Areas (Fig. 1) and 
surface water discharge into the southern 
Everglades became controlled. Water is pres- 
ently discharged into the park through spill- 
ways (called the S-12 structures) on the 
northern park boundary (Fig. 2) . Water from 
the two eastern structures is conducted into 

Natural Resources Report No. 16 

the Shark River Slough through the marsh 
and along a canal (L-67 Canal) on the east- 
em park boundary (Fig. 2). 


Hydrologic data were collected by the 
U.S. Geological Survey (USCS) and park per- 
sonnel. Water levels were measured at ref- 
erence station P-33 (Fig. 2) and at staff 
gage stations along a transect down Shark 
River Slough (Fig. 2) . Discharge was mea- 
sured periodically at S-12 water control 
structures along the northern park boundary 
and total discharge through all the struc- 
tures was then calculated by the USGS. 
Drying rate is defined as the rate of water 
level change from the highest water level 
in one month to the lowest level 3 months 
later as measured at reference station P-33 
{see Kushlan et al. 1975 for details). 

Two long-term fish trap sites used by 
the USGS from 1965-72 were reestablished. 
Monthly fish samples were taken with two 
4.6 m^ pull-up traps from November until 
the site dried in May. Our procedures fol- 
lowed those of the previous study (Kushlan 
et al. 1975) except that nets were left 
submerged in the water rather tlian exposed 
to air during the period between samples, 
and sampling was done during the day rather 
than at night. The latter change was jus- 
tified because statistical analyses of data 
collected by the USGS showed no significant 
differences were detectable in numbers, 
weights, or kinds of fish caught between 
day and night trapping. 

Nine other fish- trapping stations were 
also established along a transect down 
Shark River Slough and sampled at monthly 
intervals (Fig. 2) . All but one were 
located next to permanent staff gage sta- 

tions. Monthly fish samples along this 
transect were taken with l-m^ drop -traps or 
l-m^ throw- traps, unless a station was dry. 
Fish sampling stations also were established 
at three sites on Cape Sable (Fig. 2). =^ 

Food of Wood Storks was determined by 
obtaining regurgitation samples from nest- 
lings in colonies and from adults at feeding 
sites. Nestling storks regurgitate readily 
when handled. We found that when we used a 
helicopter to fly close to feeding adult 
storks, they would regurgitate food onto 
the ground before flying away. We then 
landed to collect the regurgitation and took 
several drop-trap or throw-trap samples of 
fish at the same place where regurgitation 
samples were obtained (Fig. 2). Thus we 
were able to compare the species and sizes 
of fish available to those taken by storks. 
The data collected at feeding sites were 
combined for analysis into three habitats: 
coastal, mangrove headwaters, and Ever- 
glades . 

Fishes collected during the study were 
preserved and later identified, counted, 
and measured for total length and dry 
weight. Aerial surveys, conducted two to 
four times per month, were used to locate 
feeding storks, follow storks from colony 
sites to feeding grounds, observe nesting 
colonies, census nests, and monitor stage 
of nesting. More detailed observations on 
the development of young, number of young 
in nests, survival of nestlings, and daily 
activity of adults were obtained by ground 
visits to the rookeries. 

In addition to other aspects of the 
Wood Stork study, a color-marking and band- 
ing program was begun to obtain information 
on movements of storks outside the park and 
to gain vital information on mortality of 
nestlings and young. 

National Park Service 

Hydrologic Conditions 

During the 1974 hydrobiological year, 
1 June 1973 through 31 May 1974, rainfall 
was 15% below normal at the Everglades ref- 
erence station but 301 above normal at the 
coastal reference station (locations in 
Fig. 2) . Figure 3 indicates the above nor- 
mal rainfall for the year on the coast was 
due primarily to high rainfall in August 
and December. December was also a month of 
above normal rainfall in the Everglades. 
Rainfall between January and May was well 
below normal at both locations, only 19% of 
normal in the Everglades, and only 39% of 
normal along the coast. Tlius, despite 
periods of high rainfall, the 1974 hydro- 
biological year was characterized, overall, 
by below normal rainfall. 

Monthly calculated water discharge 
through the control structures into the 
park fell below the minimum annually sched- 
uled amount from October through January. 
These low discharges were caused by low 
rainfall, gate manipulation, and nonaccep- 
tance by the park of makeup discharge from 
350 T 


the U.S. Corps of Engineers. The two east- 
ernmost control structures were kept closed 
for 25 days in November and December re- 
sulting in only 81% of the minimum schedule 
discharge during that period. 

Reporting the total discharge through 
all S-12 structures, as is presently cal- 
culated by the USGS (and presented in Fig. 
4), does not show differences among indi- 
vidual spillways. Such information is 
critical because flows through different 
structures could have different impacts 
on Shark River Slough. Total flow cannot 
be used to show all effects of gate manip- 
ulation, and data on discharge through 
individual structures were not available. 
In an attempt to estimate individual flows, 
semimonthly USGS measurements of flow 
through each structure were used to deter- 
mine proportional discharge at each struc- 
ture. We applied these proportions to the 
total daily discharge calculated by the 
USGS and thus obtained an estimate of dis- 
charge through each structure. This esti- 
mate is especially inaccurate for periods 
when more than one gate change occurred 









FIGURE 3. Percentage of normal rainfall each month in the Everglades and on the coast 
June 1973 to May 1974. 

Natural Resources Report No. 16 


H Minimur 

D scheduled 


Amount received 


r 1 

£ 1+0- 






r 1 





10 - 

n - 






FIGURE 4. Total discharge through S-12 spillways in 1973-74. 

between semimonthly measurements. 

Our calculated discharge through each 
S-12 structure from October through 
December 1973 is shown in Fig. 5. During 
the November- December manipulation, flow 
through S-12C ceased while flows through 
S-12A and B increased. S-12D remained 
closed. Thus, the below-minimimi discharge 
of November and December was released 
through structures A and B and presumably 

most of it did not enter the slough. 

Water level fluctuations monitored at 
the Shark River Slough gaging station P-33 
(Fig. 6) declined rapidly through November, 
rose in December, and fell at an increasing 
rate through spring. Water at P-33 reached 
ground level in April and fell to a low of 
0.9 m (3.0 ft) above mean sea level (msl) 
in May. Water level remained below ground 
surface until mid- June. 

National Park Service 

No conclusion is possible as to whether 
the short period of gate manipulation had 
any effect on the early drying rate. As 
noted above, discharge from October through 
January was below schedule, particularly so 
in November because of the manipulation 
(Fig. 4). Rainfall was also below normal 
in October and November (Fig. 3) and during 

this period water level fell rapidly. How- 
ever, the high December rainfall (Fig. 3) 
raised water levels which decreased the 
overall November- January drying rate. 
Therefore, in 1974 below schedule dis- 
charges were not in themselves enough to 
provide a very high drying rate. The early 
drying rate (November -January) at P-33 was 


I I I I . H 








dlo g 

(id I 


■ 1 1 

i i 



' I h ■ i 
I L : 1 ' 



II. I 


/5 ^p 









/^, 1^**. Jtf , i rf] 1 -f-' 



/• IJ . , , fJff A^ |J» J yfi ,jr [.• 

M ^ 'f _ " 

A' J5 3? , . 

FIGURE 5. Discharges through each S-12 spillway from September through December 1973. 
(S-12D was closed with no discharge during this period.) 

Natural Resources Report No. 16 

was 0.14 cm/ day; the late drying rate 
(January-March) was similar at 0.15 cm/ day. 

Some information also was gathered on 
changes in water levels elsewhere in the 
slough. Figure 7 shows, to the best of cur- 
rent information, the general pattern of 
drying in the southern Everglades. In 
general, the slough first dried around its 
middle and then toward the northeast and 
southwest from this area. As the dry sea- 
son progressed, these two tongues of water 
retracted until May when the remaining 
pools were northeast of Rookery Branch and 
southwest of the L-67 extension canal 
where water was continuously replenished 
through the spring by discharge through 
S-12D. This general pattern was derived 
from intermittent aerial surveys and is 
not precise in many details. More de- 

tailed information is required on the im- 
portant questions of how the slough dried. 

Two of the present study results are of ^ 
interest in respect to the analyses and 
conclusions of Kushlan et al. (1975). ^^^:^' 
First, drying rates observed are similar 
to those that, before 1962, were associated 
with failure of Wood Stork nesting. Suc- 
cessful stork nesting during the study year 
supports the conclusion that the Everglades 
system has changed in recent years. Second, 
the suggestion that early dr)''ing rate may 
be correlated with time of colony formation 
is strongly supported by the events of 1974. 
The regression derived by Kushlan et al. 
(1975) predicts that a drying rate of 0.14 
cm/ day would be associated with colony 
formation around 26 January, which is pre- 
cisely what happened (see Nesting Success) . 


FIGURE 6. Water level changes at Everglades gaging station, P-33 from June 1973 through 
June 19714. 

8 National Park Service 

November 8 

January 9 

FIGURE 7. Pattern of surficial drying in Shark River Slough, 1973-74. 

Natural Resources Report No. 16 

Movement Patterns and Feeding Areas 

Most storks arrived in the park from 
northern summering areas between early 
November and early December 1973. Storks 
first concentrated on coastal feeding sites 
on the west coast and Cape Sable (Fig. 8). 
A maximum count of 1,568 birds on 4 Decem- 
ber was nearly the exact count expected on 
aerial censuses based on estimated 1973 
production and mortality. Based on nest 
counts, the 1974 breeding population was 
2,000 storks (1,000 pairs). 

Storks continued to feed in typical 
sites along the coast through December, 
with a gradual movement into mangrove head- 
waters. Storks in breeding condition were 
first observed in early December, but the 
unusually heavy rainfall and subsequent 
water level rise of that month were cor- 
related with a dispersal of birds that 
apparently delayed rookery formation. Only 
890 storks were counted on the early 
January survey and these were widely dis- 
persed. By 24 January, storks had again 
concentrated at coastal feeding sites and 
two of the three rookeries were formed. 

During February and March, high numbers 
of storks were feeding in lower Conserva- 
tion Area 3A, with lesser numbers at feed- 
ing locations in mangrove headwaters and 
in Shark River Slough. Their preference 
for feeding in the Conservation Area 
rather than in the mangroves is not clear. 
At that time, we sampled higher densities 
of important prey species in drying creeks 
along mangrove fringes, particularly near 
Rookery Branch, than in Conservation Area 
3A, and storks using the Conservation Area 
had to fly 50 km, many spending the night 
at the feeding area. 

By early April, Conservation Area 3 was 
almost completely dry, and the major con- 

centration of feeding storks shifted to the 
drying lower end of the Shark River Slough 
and adjacent mangrove fringes. A drying - 
pool in Upper Shark River Slough was also 
used by feeding storks. 

By May nearly all of the Shark River 
Slough and most of the mangrove areas had 
dried. Few feeding locations were appar- 
ently available in the park and storks made 
extremely long flights to feeding grounds 
well north of the park. Adult storks only 
returned to rookeries to feed young, re- 
maining there an hour or less. Mid-May 
surveys located feeding storks in Conserva- 
tion Area 2 (Fig. 1), 130 km north of the 
colonies, and lesser numbers on drying 
ponds and in mangrove swamps northeast of 
Everglades City, 75 km from the colonies. 

Nesting Success 

Three nesting colonies were established 
in the park in 1974 (Fig. 2). Lane River 
and East River rookeries formed between 15 
and 24 January; Madeira Rookery formed 
during the first week of February. There 
were approximately 250 nests at Madeira 
(50 of which were deserted during incuba- 
tion in late April), 150 at East River, 
and 600 at Lane River. A total of approxi- 
mately 950 nests produced young. No storks 
nested on Cuthbert Island, an island often 
used by nesting storks in past years. East 
River Rooker>^ has been much larger in past 
years. Storks and egrets, which park files 
show formerly nested at East River, are 
apparently shifting to the growing Lane 
River site, and it is possible that a 
similar shift has occurred in recent years 
from Cuthbert to Madeira. 

Madeira was the most synchronized rook- 
ery', forming last and finishing by 27 May. 
Both East River and Lane River rookeries 


National Park Service 

were relatively unsynchroni zed , with egg- 
laying at Lane River occurring over a 5- 
week period. It is not known why the syn- 
chrony differed in the colonies, but it 
may be the result of late-nesting birds 
merely being attracted to the largest 
colony. Fledging young remained in East 
River until the second week of June and 
in Lane River until the third week of 

Counts early in the nesting season 
showed an average of 2.5 young per nest. 
Survival during the first month after 
hatching was not documented. Exact counts 
of older young that died prior to fledging 
showed that these losses were 0.25 young 
per nest at Madeira, 0.20 young per nest 
at East River, and 0.18 young per nest at 
Lane River. Overall, adults fledged about 
2.0 young per nest, producing about 1,900 

This production is a substantial re- 
cruitment to the previously decreasing 
population of Everglades Wood Storks. If 
nesting is considered at least marginally 
successful when production is sufficient 
to permit a stable population from one 
nesting season to the nest, 1974 was the 
first successful nesting in 7 years and 
only the third successful season since 
the institution of surface water control 
in 1962. A question of concern is the 
mortality of these young over the 6 months 
following fledging. If late nesting and 
a presumably lowered food availability 
during the last month of nesting resulted 
in the production of suboptimally nour- 
ished young, mortality could be severe. 
If normal mortality occurs, presently 
estimated to be 40% during the first year 
(Palmer 1962) , aerial censuses of birds 
returning in the winter of 1974-75 should 

locate about 2,300 birds. A much lower 
return would suggest higher post -fledging 

Three hundred nestlings were banded, 
tagged with red nasal saddles, and sprayed 
on the wing with yellow-gold picric acid 
dye. Five reports of marked birds were 
received later. The two most distant 
recoveries were one marked juvenile found 
dead in early June at Sebastian Inlet near 
Pelican Island National Wildlife Refuge in 
Brevard County, Florida, and one found 
alive on 1 July near Ft. Myers, Florida. 

Everglades Wood Storks nested later in 
1974 than any previous successful year and 
near the limit of having sufficient time 
to complete the nesting cycle. Nesting 
was completed successfully because of a 
late start of the wet season and an un- 
usually slow rise in water level. 

During most of the season, nestling 
storks appeared to grow normally and in 
many cases the usual brood of two to three 
birds survived until late in the nestling 
period. By 14 May, however, 7 nestlings 
33-39 days old averaged 35-6 below the ex- 
pected weight from Kahl's (1962) study, 
20 nestlings 40-44 days old averaged 11% 
below expected weight, and 6 nestlings 
45-47 days old averaged 20% below expected 
weight. It is believed nestling weights 
were subnormal late in the season because 
of lowered availability of food supplies 
as water levels began to rise. 

The discovery that storks continued to 
forage and feed young despite slowly ris- 
ing water levels in late May and June, 
and that they traveled over 130 km from 
the colony to find remaining pools of 
water is extremely important. A late 
start of the rainy season and the advanced 
stage of nesting in early summer made 

Natural Resources Report No. 16 


November 8 

FIGURE 8. Feeding sites of Everglades Wood Storks in 1973-74. Each dot represents ^ 
five storks . 


National Park Service 

February 21 & 22 

March 21 

Natural Resources Report No, 16 


continued nesting possible. The eventual 
success of nesting proves that sufficient 
food can still be produced in the highly- 
altered south Florida environment to permit 
reproduction. Prior to this study, it was 
suspected that this might not be the case. 
Perhaps it is not so much the food supply 
that runs out at the end of the dry season, 
but time in which to locate it. It remains 
probable, however, that the overall fish 
production in the southern Florida eco- 
system had declined in recent years due, 
at least, to loss of habitat. Before 
1962, successful stork nesting occurred 
in high water years, suggesting that there 
was a greater total biomass of fish avail- 
able then than now, assuming that similar 
fish densities were necessary for stork 
feeding under such high water conditions. 

Fish Populations 

Reestablishment of previous USGS mon- 
itoring stations provided an opportunity 
to compare fish densities in 1974 to those 
from the period of 1965-72. Figure 9 
shows changes in fish populations in the 
Everglades as indexed by the two monitor- 
ing sites. Populations of fish and prawns 
were generally high in 1973-74. Maximum 
recorded fish density was higher than that 
of 1967, the last year when Wood Storks 
nested successfully. 

Fish densities in other parts of the 
slough and on Cape Sable were measured 
monthly at the permanent drop- trap sta- 
tions shown in Fig. 2. Two trappings per 
month at Cape Sable sites were too vari- 
able to statistically detect small changes 
in density. With a mean density of 16.8 
fish/m^, there were no significant differ- 
ences in fish density at the three Cape 
Sable sites (analysis of variance, P>0.05). 

Similarly, no significant differences were 
discerned among trappings in Shark River 
Slough during November and December (analy- 
sis of variance, P>0.05). This indicates 
that within the resolution of the trapping ; 
technique fish were distributed homogeneous- 
ly throughout the slough prior to January, 
except for station 171^ which had higher 
densities (studentized range test, P<0.05), 
probably because of its location on the 
edge of a stream. Excluding station 173^, 
the average density in the slough during 
this period was 6.9 fish/m^. 

Changes began to appear in the slough in 
January (Fig. 10). Average Januar)' fish 
density of 9.0 fish/m^ in the upper slough 
was similar to previous densities. However, 
density in the lower slough had decreased 
significantly to 3.6 fish/m^ (comparisons 
by t-test, P<0.05). This difference con- 
tinued through February. No samples were 
taken in March. By April, the lower slough 
had dried, and densities at the t^vo inter- 
mediate stations in the upper slough were 
higher than previous levels while the other 
two still -flooded stations were unchanged. 
The high mean values of 71.0 and 35.3 fish/ 
m^ for the intermediate stations could only 
have come about by movement of fish into 
these trap sites. All trap sites were dr>^ 
in May. 

The sampling program in the Shark River 
Slough detected increases in two stations 
late in the dry season, which reflect fish 
movement, but was not sufficiently sensi- 
tive to detect changes at lower fish den- 
sity. This shortcoming and lack of sam- 
pling in March preclude any quantitative 
statement on fish movements in the slough. 
In addition, it should be noted that the 
pattern of drying in the slough (Fig. 7) 
suggests that fish movement is much more 


National Park Service 

- cL^y] 

1965 ' 1966 









FIGURE 9. Long-term record of fish populations in Shark River Slough showing how 
conditions during the current study 1973-74 compared to previous years. 

complex than would be shown adequately by 
a single transect. The program (which, be- 
cause of few previous attempts at this sort 
of study, must be considered a preliminary 
one) did show that the methodology used can 
detect tenfold changes in density. However, 
detecting differences among the low densi- 
ties of fish that prevail in the Everglades 
for most of the year will require the de- 

velopment of more sensitive sampling pro- 
cedures . 

Prey Density 

There was a sharp contrast between fish 
densities at permanent sampling sites and 
at locations where storks actually fed. 
Both in the slough and on Cape Sable fish 
densities were higher at locations where 

Natural Resources Report No. 16 




30 n 








T 1 April 

% 70- 



^ 60n 



/ \ ' 


/ \ 






' \ 



L-67 2 5 7 9 P-33 1^ l6 17|- 
FIGURE 10. Fish density at stations on northeast to southeast transect down Shark River 
Slough (Fig. 2). Means are plotted. Ranges are indicated by vertical lines. 

Storks fed (t-tests, P<0.u3). Along the 
coast, average density was 40 fish/m^ as 
contrasted with 16.8 fish/m^ at permanent 
trap sites. In the Everglades, average 
density at feeding sites was 141 fish/m^ 
as contrasted with 10.3 fish/m^ at the 
permanent sites. These data of fish den- 
sity quantify and extend to the Everglades 
and coastal habitats Kahl's (1964) finding 

that storks in the Big Cypress Swamp feed 
on high densities of fish and, furthermore, 
show that storks foraged in locations where 
fish densities were relatively high. 

A notable exception occurred in the 
northern Everglades. Fish densities at 
feeding sites in Conservation Area 3iA were 
lower than in Everglades marshes within 
the park (means 12.2 vs. 144.5 fish/m^) . 


National Park Service 

Despite this, there was heavy use of Area 
3A by storks. More information from the 
Conservation Area is essential, especially 
since storks attracted to this area in 
March were passing up high concentrations 
of fish closer to the colony. Although 
there were no significant differences in 
the mean fish densities among coastal, man- 
grove, and Everglades feeding sites (t-test, 
P>0.05), the trend of increasing mean den- 
sity 48.8, 82.9, and 144.5 fish/m^, respec- 
tively, suggests that storks fed on higher 
densities later in the season. 

Food studies 

Little was known previously about food 
consumed by Wood Storks in southern Florida 
or elsewhere. Kahl (1963), who summarized 
all existing information on food of Wood 
Storks, reported on stomach contents of 
seven southern Florida storks that had been 
examined, including two he collected in the 
Big Cypress, one found dead in the Big 
Cypress, and four collected at Cape Sable 
Gator Lake in 1924. All seven contained 
only fish. Thus, other than four 50-year 
old specimens from Gator Lake, no data was 
available on food of Everglades Wood Storks. 
Previously, our understanding of foraging 
relationships of Everglades storks was 
extrapolated from Kahl's study. Samples of 
adult and nestling regurgitation and data 
on fish availability collected in this 
study provide important information on this 
previously little known aspect of stork 

Table 1 lists the prey taken by Wood 
Storks in three foraging areas and food 
brought to nestlings at two colonies. (All 
scientific names of prey are also shown in 
the table.) Fish comprised practically all 
of the Wood Stork's diet in extreme south- 

em Florida. Overall, 27 species of fish 
were consumed by Wood Storks. However, 
only a relatively few kinds of fish made 
up most of the total. Flagfish, sailfin 
mollies, marsh killifish, and the combined 
species of sunfish accounted for 83% of the 
individuals and 721 of the biomass. One 
additional species, the yellow bullhead 
which made up less than 21 of the individ- 
uals, comprised 12% of the biomass. Thus, 
these five groups of fish made up 85% of 
the number and 84% of the biomass of fish 
consumed by Wood Storks. 

The only vertebrates other than fishes 
were one red- spotted newt, three tadpoles, 
and one adult frog. Seven freshwater 
prawns were found among the 3,000 food 
items analyzed. This is surprising in 
view of the extemely high densities 
attained by this minnow-sized crustacean, 
as can be seen by comparing the two graphs 
of Fig. 9, one of which includes the 
prawn. Even though densities of prawns 
at stork feeding sites reached 1,242 
prawns /m^, they were not taken in any 
quantity. It is highly probable that 
most of the prawns were ingested secon- 
darily by storks because prawns are also 
eaten by fish, especially sunfish, and 
some of the specimens were found in the 
mouth of such fish disgorged by storks. 

Collections of regurgitation from nest- 
ling storks were made at Lane River Rook- 
ery on 17-18 April and at Madeira Rookery 
on 24 April. Food brought to nestlings 
was similar to the two colony sites stud- 
ied (Table 1), with two notable exceptions. 
Sailfin mollies were more important at 
Lane River. There was a significant dif- 
ference in the size of food brought to 
young at the two rookeries (t-test, P<0.05). 
Fish from Madeira Rookery averaged 4.1 cm 

Natural Resources Report No. 16 

















































. <r 













































-H 00 .-. 

o ^ o 

00 r- .— o -H •-< -^ 

— O — ' -H 

in 00 -T 

— < 00 — 

00 O^ 00 >^ >-• 

o o o o 

O \D O 

o o o 

o — ' 

i-H r^ O 
■^ ■-" 00 

O •-' r^ .-• 

t^ 00 lH 

,— I 1— < rO Tl- (N] hO 

00 rO u-i 

.-I ul 

00 00 ro 

■g C 

:3^ s ^ ?) to 

'^ 63 to W 

0) O 3 3 3 

;5 J- t-i f-i «. 

^ -g -^ -^ -^ 


E; o ti, u. t 


(i) Ci3 CU Q) 

^ — ^ 



J= *-> D 

— < ^ a- <u 

4) — . — 

_j E _ 

— • C3 — < 

— t- 1 CJ -O — ' — ' 


National Park Service 



■=" 100 








1 2 c i+ 6 8 10 
Size (cm) 

r - I I 

12 \\ 15 

200 - 

150 - 

^H 100 




12 h 6 8 10 12 
Size (cm) 



FIGURE 11. Size frequency distribution of fish from nestling Wood Storks at two colonies, 

compared to 5.4 cm at Lane River Rookery 
(Fig. 11). The difference was due to a 
greater proportion of larger fish species 
5-9 cm) obtained in the Lane River samples 
(Table 1) and not to differences in size 
of similar species. 

Nestlings consumed larger fish on the 

average than did adults (t-test, P<0.05). 
The differences were not great (4.6 cm vs. 
4.3 cm) and may not be meaningful. 

To determine whether there were any 
differences in the size of fish fed to 
different -aged young, regurgitation samples 
from nestlings were broken down by nestling 

Natural Resources Report No. 15 


age classes. There were no consistent dif- 
ferences in size of fish fed to young of 
various ages. In fact, the second largest 
fish found in this study, a 22-cm bass, was 
from a stork less than 20 days old. These 
results do not confirm the observation of 
Kahl (1962) that smaller fish were brought 
to younger storks. 

Prey selection by Wood Storks is pas- 
sive in that storks do not visually select 
and pursue prey but rely on nonvisual 
foraging behavior (Kahl and Peacock 1963) . 
Particular species consumed may be captured 
selectively because of availability (i.e., 
density), size, behavior, or other charac- 
teristics. The relationship between avail- 
ability and prey actually taken can be 
quantified by a selectivity (electivity) 

E = ^ - ^ 
e + a 

where e is the proportion of a species or 
length of fish among all prey eaten and a 
is the proportion of a species or length of 
fish available to foraging Wood Storks as 
determined by the samples taken at stork 
feeding sites. This index, proposed by 
Ivlev (1961) , quantifies prey preference 
by a number ranging from +1 to -1. A prey 
with an index near +1 is consumed selec- 
tively in much greater proportion than it 
is available. A prey with an index near 
zero is consumed in proportion to its 
availability. A prey with an index near 
-1 is consumed below its relative abundance 
in the environment. 

Figure 12 shows the relative abundance 
of various species of fishes, their rela- 
tive importance in the diet of Everglades 
Wood Stork, and selectivety of storks for 
each species of fish. Of the more impor- 
tant species in the diet, sunfish, marsh 

killifish, sailfin mollies, and sheepshead 
minnows were highly selected. The mosquito 
fish, abundant in feeding areas, was under- 
represented in food samples obtained. Sim- 
ilarly, bluefin and least killifish, al- 
though relatively abundant, are seldom 
taken by storks. 


B« = L. 







FIGURE 12. Relative abundance of various species 
of fish, their frequency in the Wood Stork diet, 
and the selectivity of storks for each species. 

Selectivity by storks for various 
sizes of fish can be measured in the same 
way as selection for species, by calcula- 
ting selectivity indexes for various 
length classes, as shoun in Fig. 13, for 
all samples. Selectivity was zero for 
fish about 3.5 cm in length and rose 
rapidly for larger sizes. This pattern 
was similar in all three habitats. This 
means that relative to the available sizes 
of fish. Wood Storks selectively consumed 


National Park Service 

the larger fish, generally those over 3.5 
cm. Other aspects of prey selectivity of 
Wood Storks are discussed elsewhere (Ogden 
et al. 1976). 

10 u 

FIGURE 13. Selectivity of Wood Storks for 
different sizes of prey. 


Rainfall was below normal during most 
months of the 1974 hydrobiological year. 
Discharge into Shark River Slough was 
below the scheduled minimum from October 
through January because of low rainfall, 
management diversion of discharge west of 
the slough in November, and nonacceptance 
by the park of make-up discharges. Thus, 
the 1974 hydrobiological year was one of 
low water supply to the park. However, 
high rainfall in December dispersed feed- 
ing Wood Storks and delayed the initiation 
of colony formation. The drying rate, 
which is one way of quantifying the pat- 
tern of water level decline, was rapid 
■despite the December rainfall, and the 
overall drying pattern for this year was 
similar to years before 1962 when Wood 
Stork nesting failed. Nonetheless, 1974 
turned out to be a successful one for the 
Wood Storks, strengthening the conclusion 
of Kushlan et al (1975) that successful 

nesting years after 1962 do not follow the 
historic pattern. These authors also 
derived an empirical relation between dry- 
ing rate in the Shark Slough and the time 
of colony formation using data from 1962 
through 1973. If this relation continued 
to hold in 1974, Wood Storks should have 
begun nesting around 26 January. In fact, 
storks began nesting between 15 and 24 
Janiiary. This extremely close agreement 
with the previously derived relation 
strongly supports its validity. 

Because of the relatively small amount 
of water manipulation, by closing of S-12 
structures, no conclusions are possible 
on impact of such actions on fish popula- 
tions of the Everglades. Since no signif- 
icant water manipulation took place, the 
most important contribution of the present 
program was the detailed study of the 
food ecology of Everglades Wood Storks. 
This information is basic to understanding 
the ecological relationships of this spe- 
cies and provides a step toward management 
of the Everglades ecosystem. 

The study showed fish populations were 
high in the Shark River Slough in 1974 and 
that movement of fish did occur as the 
slough dried, although details of such 
movement remain obscure. Wood storks fed 
on relatively high densities of fish 
except, apparently, in Conservation Area 
3, where much future study needs to be 
directed. The Wood Storks in the coastal, 
mangrove, and Everglades regions of south- 
em Florida ate fish almost exclusively, 
consuming 27 species. However, marsh 
killifish, sheepshead minnows, flagfish, 
sunfish, and yellow bullhead made up 84% 
of the diet. Some of these species were 
highly selected by Wood Storks and con- 
sumed far in excess of their relative 

Natural Resources Report No. 16 


abundance at feeding sites, whereas others 
were less highly selected. Abundant poten- 
tial food species such as mosquitofish, 
least killifish, and freshwater prawns 
were insignificant in the stork's diet. 
Wood Storks also selectively consumed fish 
over 3.5 cm, the relatively larger fish 
and therefore larger species among those 
available. Thus, these data show that the 
Wood Stork is a highly selective feeder 
despite its nonvisual, tactile hunting 

The ecological implications and manage- 
ment applications of the results of this 
study are considerable. It is believed 
stork nesting in recent years has failed 
primarily because nesting began too late. 
Previously, no nesting had succeeded which 
began after December. Nesting began this 
year in late Janary, near the latest date 
at which nesting could successfully finish 
before the rainy season. The attempt to 
initiate early nesting by water management 
was inconclusive. However, nesting would 
probably have begun earlier if water levels 
liad not risen in December. 

It is apparent from this year's data 
that the general rate of water decline at 
the P-33 reference gaging station does 
correlate with time of colony formation, 
but why it does is not clear. It appears 
that storks were not feeding in the area 
which is most probably influenced directly 
by small changes in the discharges at the 
S-12 structures during the early drying 
period and that less than scheduled water 
deliveries did not in themselves provide 
a rapid drying rate this year. It is 
possible that relatively small manipulation 
of dry season discharge may have little 
impact. However, it remains unknown how 
S-12 discharges affect water levels eitlicr 

at P-33 or elsewhere in the slough. As 
noted by Kushlan et al. (1975), the value 
of P-33 may be as an index of the pattern 
of drying over the entire region, including 
areas where Wood Storks are feeding during 
colony formation. Thus, water level rela- 
tions at P-33 are probably a valid indica- 
tion of water conditions in relation to 
Wood Storks, but the relation between P-33 
and S-12 discharge remains unquantified 
in any way applicable to management. It 
is reasonable to suspect that high dis- 
charges in the drying season might prevent 
rapid drying. Thus it remains reasonable 
that manipulation of early dry season dis- 
charges may indeed affect drying. However, 
without more detailed information on the 
amount of discharge through each structure 
and the impact of such discharge on water 
levels throughout the slough, it is not 
possible to predict the effect of such 

Since Wood Storks consume predominantly 
a small array of species, it is the produc- 
tion and availability of these few species, 
as opposed to the availability of the whole 
array of species, which probably determines 
whether sufficient food is available for 
successful nesting. In addition, storks 
consume primarily larger fish, the ages 
of which are not presently known. This 
information is important because if the 
fish are the result of production during 
the current hydrobiological year, the 
biological and physical conditions which 
control their availability may be relative- 
ly straightforward. If they are in fact 
over one year old and have survived a 
previous dry season, the situation becomes 
more complex. Unfortunately, little is 
known about the productivity, gro\\rth, or 
ecology of these species or of the aquatic 


National Park Service 

communities of which they are a part. 

The general nonpredictability of suc- 
cessful Wood Stork nesting after 1962 con- 
tinued. It is particularly notable that 
successful years prior to 1962 were wet 
years while the 3 successful years after 
1962 were dry years. However, other recent 
dry years and also recent high -water years 
were unsuccessful. The breakdown of pre- 
dictability coincides with the institution 
of water control, although other concurrent 
environmental changes are undoubtedly in 
part responsible. Both the analysis of 
such changes and the proper management of 
the Everglades ecosystem are confounded by 
the delivery of surface water discharge to 
the western part of Shark River Slough 
rather than through its natural drainage 
east of the park. It seems highly probable 
that the breakdown of the historical pat- 
tern of wading bird nesting may only be 
rectified if the historic pattern of water 
delivery to the entire slough is again 


This project provided a beginning toward 
an urgently needed and long overdue study 
of the Shark Slough ecosystem. We strongly 
recommend that a unified research program 
be initiated to continue investigating the 
biological and hydrological dynamics of 
this vital but failing ecosystem, the cen- 
tral feature of Everglades National Park. 
We specifically recommend the following 
program as the first stage in the needed 
studies . 

Hydrology -Data on monthly discharge 
through the S-12 structures should include 
daily discharge for each structure as a 
necessary requisite of future management 

decisions regarding this discharge. 

A program to determine the pattern of 
drying in Shark River Slough and adjacent 
areas of importance to wading birds should 
be established. 

Fish Studies --k program should be under- 
taken to determine the distribution, com- 
munity structure, breeding cycles, longev- 
ity, movement, food, and productivity of 
fish populations in the southern Everglades. 
A necessary first step in this program 
must be further refinement of the sampling 
technique used to measure low densities of 
fish. Studies of the abundance and produc- 
tivity of fish in Shark River Slough should 
then be contrasted with similar information 
from the Conservation Areas and eastern 
Everglades to determine the impact of water 
management on fish populations and their 
availability to higher levels of the food 
chain. These studies should be part of an 
overall effort to develop a systems model 
of the Everglades region which will provide 
the basis for management of these ecosys- 

Studies should also be undertaken to 
determine productivity and the environmen- 
tal determinates of prey availability in 
the estuarine habitats which serve as the 
primary feeding site of Wood Storks in 
eany winter. 

Sampling should be continued at the 
two previous USGS trap sites reestablished 
this year in order to perpetuate this long- 
term record of fish populations in Shark 
River Slough. 

Wood Stork Studies- -The present study 
obtained information on the distribution, 
movements, food, and prey selectivity of 
the Wood Stork in the Everglades. Studies 
now should be undertaken to examine more 
closely its nesting biology, foraging 

Natural Resources Report No. 16 


behavior, and parental care in the Ever- 
glades while expanding the study of its 
ecology and biology to other parts of its 
range so that the environmental factors 
that lead to successful production in dif- 
ferent habitats can be compared and gen- 
eral patterns uncovered. In addition, 
banding and marking programs in the park 
and elsewhere should be continued to pro- 
vide information on mortality and distri- 
bution in the nonbreeding season. 

Wading bird studies should be expanded 
to include more precise information on 
the movements, food habits, and popula- 
tions of other wading bird species. Sur- 
veys should give better coverage of impor- 
tant feeding areas for wading birds out- 
side the boundaries of the park, especial- 
ly in the Big Cypress Swamp, the Conserva- 
tion Areas, and eastern boundary area. 

Alligatov 5twdies-- Preliminary studies 
(Kushlan 1974, 1976) have shown the impor- 
tance of this species and its role in pro- 
viding foraging habitat for wading birds 
and survival ponds for fish. A program 
should be begun to study the ecology of 
the American alligator in the southern 
Everglades, specifically in relation to 
water levels, fish populations, and wading 
bird predation. 

Future Considerations- -Hydrobiological 
research in the southern Everglades should 
have as its ultimate goal determining 
whether the current discharge structures 
and discharge schedule adequately repro- 
duce the natural hydrologic cycle in the 
Shark Slough watershed and, if so, whether 
this has maintained the natural biological 
system. An effort should be undertaken 
to determine if surface flow should be 
restored to the entire Shark Slough basin. 


National Park Service 


IVLEV, V. S. 1961. Experimental Ecology of the Feeding of Fishes. Yale Univ. Press, 

New Haven, Conn. 302 p. 
KAHL, M. P. 1962. Bioenergetic of growth in nestling Wood Storks. Condor 64:169-183. 
. 1963. Food ecology of the Wood Stork in Florida: A study of behavioral and 

physiological adaptations to seasonal drought. Ph.D. Thesis, Univ. Georgia, Athens. 
. 1964. Food ecology of the Wood Stork {Myateria amevieand) in Florida. Eaol. 

Monogr. 34:97-117. 
, and L. J. PEACOCK. 1963. The bill-snap reflex: A feeding mechanism in the 

American Wood Stork. Nature 199:505-506. 
KUSHLAN, J. A. 1974. Observations on the role of the American Alligator {Alligator 

mississippiensis) in the southern Florida wetlands. Copeia 1974:993-996. 

. 1976. Wading bird predation in a seasonally fluctuating pond. Auk 93:464-476. 

, J. C. OGDEN, and A. L. HIGER. 1975. Relation of variation of water level and fish 

availability to Wood Stork reproduction in the southern Everglades, Florida. U.S. 

Geological Survey, Tallahassee, Florida. Open-file Rept. 75-434. 56 p. 
OGDEN, J. C, J. A. KUSHLAN, and J. T. TILMANT. 1976. Prey selectivity by Everglades 

Wood Storks. Condor 78:324-330. 
PALMER, R. S. 1962. Handbook of North American Birds, Vol 1. Yale Univ. Press, New 

Haven, Conn. 


This study was supported by the U.S. National Park Service. 
We thank Dr. William B. Robertson, Jr., Gary E. Davis, and L. 
Lee Purkerson for comments on and assistance in setting up the 
study. William Robertson, Frank Nix, Gary Davis, Roy Johnson, 
Thomas Schmidt, Ronald Hermance, and Charles Hilsenbeck reviewed 
and commented on the report. Marilyn Kushlan drew the figures. 
We also thank the Ranger Division of Everglades National Park, 
particularly Mike Warren, Fred Dayhoff, Stan Robbins, and 
Fred Whitehead, for assistance. 

Natural Resources Report No. 16